Introduction

The CDC ChildVaxView tool provides state-level estimates of vaccination coverage among children under 3 years of age, by birth year. These rates are modeled using Kaplan-Meier survival analysis after aggregating multiple survey years by birth year of respondent. In our simulation, we also use the National Immunization Survey (NIS) data; however, we use raw survey responses (after accounting for survey design). Here, we compare the vaccination coverage across the two sources.

Note that the final year of CDC data (birth cohort 2021 or survey year 2024) is considered preliminary and is likely to change.

library(tidyverse)
library(here)
library(geofacet)
library(DT)

vaccine_df <- readRDS(here("data", "vaccine_coverage.RDS"))
cdc_raw <- read_csv(here("data_raw", "Vaccination_Coverage_among_Young_Children__0-35_Months__20250401.csv")) |>
    janitor::clean_names()

print_df <- left_join(
    vaccine_df |>
        transmute(
            year,
            abbrev,
            vaccine,
            nis_estimate = sprintf("%0.1f (%0.1f to %0.1f)",
                round(estimate * 100, 1),
                round(lower_ci * 100, 1),
                round(upper_ci * 100, 1)
            )
        ),
    cdc_raw |>
        filter(geography %in% state.name,
            vaccine %in% c("≥1 Dose MMR", "DTaP", "Polio"),
            dimension_type == "Age",
            dimension == "35 Months",
            birth_year_birth_cohort %in% as.character(2001:2021)) |>
        filter((vaccine == "DTaP" & dose == "≥4 Doses") |
            (vaccine == "Polio" & dose == "≥3 Doses") |
            (vaccine == "≥1 Dose MMR")) |>
        left_join(
            tibble(geography = state.name,
                abbrev = state.abb)
        ) |>
        transmute(
            abbrev,
            year = as.numeric(birth_year_birth_cohort) - 2,
            vaccine = case_when(
                vaccine == "≥1 Dose MMR" ~ "mmr",
                vaccine == "DTaP" ~ "tdap",
                vaccine == "Polio" ~ "polio"
            ),
            cdc_estimate = sprintf("%0.1f (%s)", estimate_percent, x95_percent_ci_percent)
        ) 
) |>
        arrange(vaccine, abbrev, year) 
plot_df <- bind_rows(
    print_df |>
        transmute(
            year,
            abbrev,
            vaccine,
            source = "Raw NIS data",
            estimate = as.numeric(str_split_i(nis_estimate, " \\(", 1)),
            lower = as.numeric(str_split_i(nis_estimate, " \\(| to ", 2)),
            upper = as.numeric(str_split_i(nis_estimate, " to |\\)", 2))
        ),
    print_df |>
        transmute(
            year,
            abbrev,
            vaccine,
            source = "CDCVaxView",
            estimate = as.numeric(str_split_i(cdc_estimate, " \\(", 1)),
            lower = as.numeric(str_split_i(cdc_estimate, " \\(| to ", 2)),
            upper = as.numeric(str_split_i(cdc_estimate, " to |\\)", 2))
        )
)

MMR

ggplot(plot_df |>
    filter(vaccine == "mmr"),
aes(x = year,
    y = estimate,
    ymin = lower,
    ymax = upper,
    color = source,
    group = source, fill = source)) +
    geom_ribbon(color = NA, alpha = .2) +
    geom_line() + facet_geo(~abbrev) +
    theme_bw() +
    scale_fill_brewer("Data source", palette = "Set1") +
    scale_color_brewer("Data source", palette = "Set1") +
    scale_y_continuous("MMR vaccination rate at 35 months") +
    scale_x_continuous(NULL)

print_df |>
    filter(vaccine == "mmr") |>
    select(-vaccine) |> 
    DT::datatable(
        colnames = c(
            "Year",
            "State",
            "NIS Estimate",
            "CDC Estimate"
        ),
        rownames = FALSE,
        filter = list(
            position = "top"
        ),
        options = list(
            columnDefs = list(list(targets = c(2, 3),
                searchable = FALSE)),
            paging = TRUE,
            searching = TRUE,
            pageLength = 20)
    )

TDaP

ggplot(plot_df |>
    filter(vaccine == "tdap"),
aes(x = year,
    y = estimate,
    ymin = lower,
    ymax = upper,
    color = source,
    group = source, fill = source)) +
    geom_ribbon(color = NA, alpha = .2) +
    geom_line() + facet_geo(~abbrev) +
    theme_bw() +
    scale_fill_brewer("Data source", palette = "Set1") +
    scale_color_brewer("Data source", palette = "Set1") +
    scale_y_continuous("TDaP vaccination rate at 35 months") +
    scale_x_continuous(NULL)

print_df |>
    filter(vaccine == "tdap") |>
    select(-vaccine) |> 
    DT::datatable(
        colnames = c(
            "Year",
            "State",
            "NIS Estimate",
            "CDC Estimate"
        ),
        rownames = FALSE,
        filter = list(
            position = "top"
        ),
        options = list(
            columnDefs = list(list(targets = c(2, 3),
                searchable = FALSE)),
            paging = TRUE,
            searching = TRUE,
            pageLength = 20)
    )

Polio

ggplot(plot_df |>
    filter(vaccine == "polio"),
aes(x = year,
    y = estimate,
    ymin = lower,
    ymax = upper,
    color = source,
    group = source, fill = source)) +
    geom_ribbon(color = NA, alpha = .2) +
    geom_line() + facet_geo(~abbrev) +
    theme_bw() +
    scale_fill_brewer("Data source", palette = "Set1") +
    scale_color_brewer("Data source", palette = "Set1") +
    scale_y_continuous("Polio vaccination rate at 35 months") +
    scale_x_continuous(NULL)

print_df |>
    filter(vaccine == "polio") |>
    select(-vaccine) |> 
    DT::datatable(
        colnames = c(
            "Year",
            "State",
            "NIS Estimate",
            "CDC Estimate"
        ),
        rownames = FALSE,
        filter = list(
            position = "top"
        ),
        options = list(
            columnDefs = list(list(targets = c(2, 3),
                searchable = FALSE)),
            paging = TRUE,
            searching = TRUE,
            pageLength = 20)
    )